Because most philosophies that frown on reproduction don't survive.

Thursday, July 06, 2017

Pricing for the Last Mile


We have frequent conversations at work about to what extent Amazon and eCommerce in general will take over our segment of the retail world. Such conversations have been had even more widely over the last few weeks, with the announcement that Amazon is seeking to acquire grocery specialty retailer Whole Foods, so I thought it might be of some general interest for me to lay out some of the pricing considerations which define which product categories are ripe for online conversion and which aren't. Or, to be more precise, which products can be sold online and delivered for a price which is similar to or cheaper than in-store pricing that people are used to.

It helps to think back to how Amazon got its start. Fifteen years ago, Amazon as primarily a place for buying books, CDs, and DVDs. It was very successful at this, and it offered prices significantly below the prices that people normally payed for these products in traditional bookstores. Why?

Selling books is an inventory game. Sure, it's easy for a store to stock the latest mass market best seller. These are the books you'll see in your supermarket or a general merchandise store like Target, often marked down to 20-25% off cover price. Retailers can sell best sellers at low prices because they know that lots of people walking through the store that week will want to buy that specific book. The problem for booksellers is dealing with the economics of all the other titles that aren't best sellers. Go into a bookstore and you see shelves holding tens of thousands of books, in a big one, hundreds of thousands. You go to the Mystery section and pick up the particular Sayers novel you wanted to read. They have three paperback copies. Think about the inventory investment that makes that possible. Your bookstore stocked thousands of Mystery novels, including three copies of the particular novel you turned out to want. When was the last time someone wanted to buy Gaudy Night at that one location? When is the next time someone will? That book may have been sitting on the shelf, waiting for you, for six months or a year. Bookstores have to make these inventory investments in books that will sit on the shelf a long time, because nothing turns a dedicated bookworm off like going into a bookstore and not finding the book he wants. What's he going to do? Go find another store which does stock it. And these days, that store is probably Amazon. Let the book buyer become loyal to Amazon, and you've just shrunk your business.

So investing in inventory is important for sellers of books and media, but it's also very expensive to buy a bunch of inventory which will sit around for months before being sold. In order to allow bookselling to be economically viable, publishers have been required to price fairly healthy margins into the cover price of a book. In order for your bookstore to have that book sitting around waiting for you to want it (something which in retail terms is called 'slow turning inventory') they have to make a pretty healthy cut of the cover price when it sells.

This is why current best sellers have traditionally been discounted. If you know that you can sell a lot of copies of a book very quickly after bringing it in to the store, it's okay to sell it for a lower price. If, on the other hand, it might be months before you sell a book, you need to make more profit on each copy that you sell in order to make up the expense of stocking it for so long. A standard markup from publisher to bookseller was 40-50%, so if a bookstore sold a best seller at 20% off, it was essentially agreeing to give away half the profits on each copy because it knew that it wasn't going to have to sit on the inventory for very long.

Amazon was able to upend this model because it didn't have to stock lots of locations. Barnes & Noble (which itself was able to have a larger inventory by having notoriously large locations which people were willing to travel a little bit further to reach) has 640 locations around the US. They might stock 2-3 copies of any given midlist novel at each location, but a lot of those copies might sit around for months waiting for someone to come along and take them home. Books that turned even slower, Barnes & Noble might not stock at all. If only ten people around the county were going to buy a book in a given month, it doesn't make sense for B&N to stock a copy at each of their 640 locations on the chance that one of those ten people might walk into that location looking for it.

By working mail order, Amazon was able to solve this inventory problem. They could stock twenty copies in one warehouse and meet the demand of all ten of those customers each month, with an inventory turn rate of one every two months. To meet the same demand and be sure to turn no customer away, B&N would have had to stock perhaps two copies in each of their 640 stores. They'd have to own 1,280 copies instead of 20 and would have a turn rate of one every ten years instead of one every two months.

To get product to customers, Amazon had to mail books to people. This was and remains a huge expense for Amazon. However, because Amazon didn't have to maintain hundreds or thousands of stores, and was able to maintain much more productive inventory, Amazon was able to afford the cost of mailing books to consumers and still offer prices much lower than a brick-and-mortar bookstore might.

To look at an example: I checked the current top Mystery novel page on Amazon and found a novel called Human Acts. Its cover price is $22.00. Amazon is selling it for $14.95, a discount of 32%. If we assume that Amazon gets the book at a 50% discount, it's cost is $11. That leaves them $3.95 to cover the cost of shipping the book, which has a shipping weight of 0.75lbs Given Amazon's economies of scale, they can pull that off and break even on the deal. However, it's important to realize that this works because books are relatively expensive, offer high retailer profit margins, and are comparatively expensive per pound.

Now let's look at a product which you might buy at the grocery store. A gallon of milk might sell for $2.49. It weights just over 8lb. The retailer's profit margin on it is perhaps 10%. There is no way that the grocery store can deliver that milk to you for a cost of less than $0.25, so delivering it to you would certainly be done at a loss.

Milk is often pointed to as a loss leader, though. Let's look at something with better margins. Cat litter is a product people buy regularly since it is consumable. It doesn't have to be refrigerated. A 20lb jug costs $7.95 at my grocery store, and retailer profit margins are close to 50%. If we assume that 50% margin, the profit per unit of a jug of cat litter might be about the same $4 as the book I mentioned above. The problem is that it weights 20lb instead of 0.75lb. There's no way you can profitably mail that.

Why does Amazon not have an advantage over the grocery store here?

First, there is not inventory turn advantage. Milk and cat litter are both products that hundreds or thousands of people will want to buy every week at each grocery location. The grocery store can bring them in in large quantities and be confident that they will move the inventory out again very quickly. Amazon does not realize any logistical savings by having dozens of huge warehouses instead of thousands of individual stores, because even a single store can still move plenty of milk or cat litter. This means that the margin structure is not much different for Amazon and the grocery store on these products.

Second, Amazon has to spend shipping expenses in order to get product from their warehouse to a customer's house, whereas at a grocery store the customer transports the product from the store to their house. This means that Amazon actually has a cost of doing business that groceries and other brick-and-mortar businesses do not have. In the cases that we just looked at, this is compounded by the fact that the products are very heavy.

So Amazon competes best in product categories where customers want access to a large selection of products, each one of which might be fairly slow to turn. It also does best in product categories where the retail price per pound is relatively high.

This can help to explain which categories Amazon entered first as it moved on from its original specialty in books, music and movies. Consumer electronics were an early growth area. A flat screen TV might be relatively heavy. However, a 30lb TV selling for $500 is still a much higher price per pound than an 8lb jug of milk selling for $2.49. Electronics, power tools, etc. are also durable products which are purchased less frequently, so Amazon can use it's more centralized inventory to some effect there again.

Of course, Amazon now sells everything, from books and music to Tide detergent and cat litter. However, it's important to notice that when it comes to grocery type items, Amazon is no longer selling based on having the lowest price. Here's a two-pack of the type of 20lb cat litter jug that I buy at the supermarket for $7.95 each. However, Amazon is selling the two-pack via Prime for $34.46. It also offers an option to buy a smaller 14lb jug for higher price of $8.31 -- but only with the requirement that you either buy at least five Prime Pantry items that can be shipped together or else pay a $5.99 shipping fee.

At one consumer packaged goods company I am familiar with, a survey of Amazon prices showed that the Amazon Price price for 20% of products was more than 25% above the MSRP for the product. This isn't because Amazon is out to cheat people, it's because on grocery and items which are easily picked up at local stores, Amazon isn't selling a wider selection or lower prices, it's selling convenience. It makes up for the high cost of getting the product that last distance from distribution point to your doorstep by charging higher prices than the stores where you'd have to transport the product home yourself.

There are grocery store type products which could make a fair amount of sense to tell online at competitive prices, but they would have to be products which either rely heavily on selection (exotic foods, fine wines, etc.) or else products which have a high price per pound (razor blades, specialty cosmetics, etc.) It will not be cheaper to buy basic grocery items at Amazon rather than in-store any time soon.

4 comments:

Brandon said...

Thinking over the non-books, non-music things I have occasionally bought from Amazon, this makes a lot of sense from the customer side, as well. They are always things that would be hard for me to get in person (unusual batteries), or things that I'd have to go through the hassle and expense of packing up and shipping myself anyway (various kinds of gifts), or both (Italian limoncello as a Christmas present). I can't conceive of any situation that I am likely to encounter in which it would make more sense to order milk by mail than just go to the nearest grocery store.

Rob Alspaugh said...

Is it safe to assume that Amazon has sussed out all the products that they can sell for profit under this analysis, and that the Next Big Thing will be based on some sort of totally new approach in a different kind of market? Or are there still people out there trying to figure out other weight/convenience shipping profits?

Darwin said...

Part of what makes the process of figuring this out is that Amazon itself is just a retailer. They make it clear to the manufacturers that sell them products that they are not willing to lose money long term on a product. If the manufacturer will not sell them the product a cost which allows Amazon to at least roughly break even while covering shipping costs, they'll turn off the SKU and not sell it.

This leaves manufacturers to decide whether they want to give Amazon a better cost than their other retailers, or give up the potential sales that come with being listed on Amazon.

Also, in the long term, Amazon is pretty clearly pursuing a strategy of being so convenient (see buying buttons for household staples, Alexa devices which pick up verbal requests to order something, etc.) that people won't worry about the price much. If they can do this, at least with some portion of the consumers out there, they will make products become economical to ship which currently aren't.

Michael said...

Darwin wrote: "Amazon itself is just a retailer"

this is only partially true. Amazon is also a services company. Their AWS backend runs the store, but is also offered to other businesses for generic cloud computing. Their store is also a platform for "Marketplace" sellers. Amazon has famously low profit margin, and I wouldn't be surprised if a substantial portion from their non-store divisions.

I found this analysis compelling regarding the Whole Foods acquisition:
https://stratechery.com/2017/amazons-new-customer/